Experimentals Record 25% Less Fatal Accidents In 2013

[Geico266]

... And so far 2014 is under 2013! Yahoo!

Nice to see some of the training issues EAA resolved with the FAA bearing fruit... in lives and aircraft saved.


According to Sport Aviation magazine, March 2014

 

[frfly172]

Great news.

 

[RotorAndWing]

I would like to see the whole data. If you dig deeper I believe you will find one reason the accident rate is down is a direct correlation with reduced flight hours.

 

[Geico266]

I would like to see the whole data. If you dig deeper I believe you will find one reason the accident rate is down is a direct correlation with reduced flight hours.

There is no official "cause and effect" conclusion made in the article. In fact, they are asking EAB builders to take a survey and try and come up with the answers. The good news is EAA is carefully looking into cause and effect, and the accident numbers are way down. 25% is a good start, no?

Seeing as how EABs are growing faster and larger as a fleet I doubt less flying hours is the cause. The reduction in accidents is in total numbers of accidents, not accident rates per hours flown.

*** Interestingly, the EAA Safety board has recommended and pushed the FAA for a "Second Pilot" during Phase 1 flight testing. This is another place where regulations have hurt the EAB safety record. Having another experienced pilot (one who has flown that model) on board is a good idea when flying off the 40 hours. Helps reduce work load and the second pilot may see issues the inexperienced pilot may not have noticed.

 

[Bob Noel]

I would like to see the whole data. If you dig deeper I believe you will find one reason the accident rate is down is a direct correlation with reduced flight hours.

Another possible reason: If fewer new aircraft were completed... this would avoid some of the problems with low-time aircraft and pilots with little time in the aircraft.

Note: I'm not say that this is the case, just offering another possible reason.

Also note: I'm happy the number of fatal accidents in lower.

 

[Geico266]

Another possible reason: If fewer new aircraft were completed... this would avoid some of the problems with low-time aircraft and pilots with little time in the aircraft.

This is where the second pilot during "Phase 1" flight testing would help. Before you HAD to fly solo, before there was little opportunity for transitional training in EAB.

Nice to see suggestions made to EAA by home builders are taken to the highest levels of the FAA for action. This new rule will save lives and airplanes.

 

[TCABM]

It will be interesting to see if part of the causal effect is attributed to a rise in E-LSA kit completions and whether the E-LSA kits have a substantially different build process than traditional EAB kits.


Sent from my iPad using Tapatalk

 

[Geico266]

Also note: I'm happy the number of fatal accidents in lower.

It is good for aviation we continue the trend.

 

[Geico266]

It will be interesting to see if part of the causal effect is attributed to a rise in E-LSA kit completions and whether the E-LSA kits have a substantially different build process than traditional EAB kits.


Sent from my iPad using Tapatalk

Having built both I can tell you the E-LSA kits must be completed exactly to the plans, where EAB you can modify want you want. Both need FAA inspection to get an AW Cert. Generally, the build process is the same though, with EABers getting a little more "creative".

Almost 1/2 of all accidents are still "Pilot Error".

Our very own Ron Wannttaja (I think he hangs out here ) wrote the article. Maybe he can comment on his article and findings.

 

[steingar]

Like other posters said, while nice the reduction in crashes is not per se meaningful. Given the lousy state of the economy I wouldn't be surprised at fewer completions or flight hours. Still, any improvement is cause for cheer.

Flight hours will be down this year for certain. Weather.

 

[yakdriver]

A lot of it has to do with the design and ease of building the newer kits on the market. The popular kits are all pretty much self jigging so it's hard not to build them straight. It is also easier to get some dual in most of the popular models like Van's, Zenith, Glassair etc. Allowing two pilots on board during the test period is a great idea. A lot of accidents are in the early phase of flight testing by builders who don't have the qualifications but insist of doing the first flight. I also wonder how many non injury accidents don't get reported. Stuff like running off the runway during a high speed taxi test and bending the airplane up then hauling it off without saying anything to anyone. Don

 

[TCABM]

Thanks, Geico. I haven't built anything yet, but am considering the RV-10, F33 Bo, and M20J airframes for my particular mission with a completion or purchase about 5 years out. The tipping point is a nice mid -90s 201 can be had for quite a bit below a -10. However, after a motor and avionics, it gets pretty close to even, $ wise.


Sent from my iPad using Tapatalk

 

[wanttaja]

Having built both I can tell you the E-LSA kits must be completed exactly to the plans, where EAB you can modify want you want. Both need FAA inspection to get an AW Cert. Generally, the build process is the same though, with EABers getting a little more "creative".

Almost 1/2 of all accidents are still "Pilot Error".

Our very own Ron Wanttaja (I think he hangs out here ) wrote the article. Maybe he can comment on his article and findings.

Sounds like my cue.

If you use the NTSB database query page and select the "Amateur-Built - Yes" option, you'll see 154 accidents in 2013, yet 219 the year before... and 237 the year before that. But the year before THAT, it was 207.

So variation occurs. But personally, when there's a 30% drop between two years, I'm more inclined to suspect some other factor than a sudden improvement in safety.

Next December, I'll start assembling the statistics relative to 2013. If safety efforts are responsible for the drop, we should see a big improvement in one or two cause factors. If it's peanut-buttered across ALL the casual factors, then something else is in play. Reduction in flight hours, change in NTSB record-keeping, etc. Remember, the "Amateur-Built" flag in the FAA records is used to indicate non-Standard airworthiness, not EX-AB.

New homebuilt registrations are down a bit. Back in 2008, we were seeing about 1300 new homebuilts added to the rolls every year, but in 2013, there were about 950.

Still, though, we *are* seeing a gradual improvement in homebuilt safety. Here's the total Experimental Amateur-Built accidents per year since the start of my database:

1998 226
1999 218
2000 221
2001 211
2002 217
2003 199
2004 187
2005 208
2006 201
2007 221
2008 213
2009 240
2010 188
2011 235
2012 213

Note that, over this period, the number of accidents is essentially flat. But in that time, roughly 1,200 homebuilts per year were added. In 1998, there were about 20,000 homebuilts on the rolls. At the end of 2013, there were about 28,000... and 7,500 additional homebuilts had been removed since 2010 due to the de-registration process! So from 20,000 to (essentially) 33,000, and the number of accidents is about the same.

As I recommend in the article, when looking at accident rates, one should take the Long View...and not get too excited (or depressed) over short-term variation.

Ron Wanttaja

 

[Geico266]

Note that, over this period, the number of accidents is essentially flat. But in that time, roughly 1,200 homebuilts per year were added. In 1998, there were about 20,000 homebuilts on the rolls. At the end of 2013, there were about 28,000... and 7,500 additional homebuilts had been removed since 2010 due to the de-registration process! So from 20,000 to (essentially) 33,000, and the number of accidents is about the same.

First and foremost, Thank you for your contributions to aviation.


If the number of accidents is flat and we are adding roughly 1,000 aircraft to the fleet a year isn't that a good improvement? Certainly, hours flown is a factor, but I would think that is flat or increasing as the number of planes is added?

 

[Geico266]

A lot of it has to do with the design and ease of building the newer kits on the market. The popular kits are all pretty much self jigging so it's hard not to build them straight. It is also easier to get some dual in most of the popular models like Van's, Zenith, Glassair etc. Allowing two pilots on board during the test period is a great idea. A lot of accidents are in the early phase of flight testing by builders who don't have the qualifications but insist of doing the first flight. I also wonder how many non injury accidents don't get reported. Stuff like running off the runway during a high speed taxi test and bending the airplane up then hauling it off without saying anything to anyone. Don

This is being addressed right now by being able to add a second pilot to Phase 1 testing. This should cut down these accidents dramatically, or double the number of fatalities.

 

[kyleb]

This is being addressed right now by being able to add a second pilot to Phase 1 testing. This should cut down these accidents dramatically, or double the number of fatalities.

This particular issue gives me heartburn. Test flying needs to be a single pilot operation. The pilot should be current and fully capable of handling the aircraft if a problem arises. If the builder doesn't meet those criteria, s/he should hire a test pilot and in the meantime get suitable transition training elsewhere. Test flights are no place for pilot familiarization and checkout flights.

Adding a second occupant doubles the risk and can also confuse PIC issues, leading to worse than necessary out ones when things do go south.

 

[Geico266]

This particular issue gives me heartburn. Test flying needs to be a single pilot operation. The pilot should be current and fully capable of handling the aircraft if a problem arises. If the builder doesn't meet those criteria, s/he should hire a test pilot and in the meantime get suitable transition training elsewhere. Test flights are no place for pilot familiarization and checkout flights.

Adding a second occupant doubles the risk and can also confuse PIC issues, leading to worse than necessary out ones when things do go south.

Good point. It could raise the risk, but in the real world I think it reduces it. The more experienced pilot / builder will be able to recognize a problem faster, and help with transition training. In any event, it looks like it will be adopted shortly.

 

[Geico266]

. Stuff like running off the runway during a high speed taxi test and bending the airplane up then hauling it off without saying anything to anyone. Don

Just to let you know Vans, and the EAA no longer recommend high speed taxies, unless it is a new design.

 

[yakdriver]

This particular issue gives me heartburn. Test flying needs to be a single pilot operation. The pilot should be current and fully capable of handling the aircraft if a problem arises. If the builder doesn't meet those criteria, s/he should hire a test pilot and in the meantime get suitable transition training elsewhere. Test flights are no place for pilot familiarization and checkout flights.

Adding a second occupant doubles the risk and can also confuse PIC issues, leading to worse than necessary out ones when things do go south.

I have done several first flights on homebuilts. I agree that the first few flights should be done solo by a highly experienced and current pilot. I don't want any distractions from the task and I will not do an initial test flight with a crowd there to watch. I usually have the owner and my good friend who is an IA and experienced in test flying himself present. Usually you can work out any problems in 5 hrs or less. It would then be nice to legally put the owner in the airplane and get him transitioned so he is able to fly off the remaining hours. How many builders want to let/pay someone else to fly off the 40hr test period? More and more builders do let an experienced pilot do the initial testing because now most planes like RVs are in the $100,000 range and they don't want to risk such a significant investment. I think this has cut down on the accident rate in the test period. Don

 

[N801BH]

First and foremost, Thank you for your contributions to aviation.


.....

Agreed 100%.... Thanks Ron for all the hard work compiling the data.....

And pilot error is a large contributing factor in fatals....

I personally know of one in 2013 and I NEVER....NEVER... NEVER... EVER... want to see that again.

 

[wanttaja]

First and foremost, Thank you for your contributions to aviation.


If the number of accidents is flat and we are adding roughly 1,000 aircraft to the fleet a year isn't that a good improvement? Certainly, hours flown is a factor, but I would think that is flat or increasing as the number of planes is added?

Well, there were 2,296 total aircraft accidents in 1998, and 1,804 in 2012. So claiming the homebuilt rate has improved is a bit tougher. When I get home tonight, I'll do a fleet-rate comparison between the years.

Ron Wanttaja

 

[Jimmy cooper]

Could be the economy, fewer people flying small aircraft. I always thought the cause was usually pilot error. It sure seems this way if you read the accident reports. Things like losing control under IFR conditions is a big one, running out of fuel is popular, missed approach screw ups, on and on.

 

[Neal Howard]

Just from my casual observations, it seems that the number of hours of flying that the average Vans RV owner flies annually, have actually picked up a little bit over the past two years. At least this has been noticeable amongst all the RV'ers I know here in the north TX / southern OK region, and by conversations I've had with the folks I hung out with in HBC at Airventure in 2012 and 2013 (record numbers of RVs showed up in HBC/HBP those years, of course Van's 40th Anniversary in 2012 had a lot to do with that).

 

[wanttaja]

Just from my casual observations, it seems that the number of hours of flying that the average Vans RV owner flies annually, have actually picked up a little bit over the past two years. At least this has been noticeable amongst all the RV'ers I know here in the north TX / southern OK region, and by conversations I've had with the folks I hung out with in HBC at Airventure in 2012 and 2013 (record numbers of RVs showed up in HBC/HBP those years, of course Van's 40th Anniversary in 2012 had a lot to do with that).

Might want to inform the FAA. If you fly a 172, they assume you fly 200 hours per year. If you complete your RV-8 and sell your 172, they assume you now fly only 29 hours per year.

That 29 hours per year is an overall figure for all homebuilts, based on the assumption that X% of the registered homebuilts were inactive. After getting rid of 7,500 homebuilts between October 2010 to December 2013, one would hope they change that assumption....

Ron Wanttaja

 

[Velocity173]

2014 has been a good start because the weather has been crappy. I've hardly flown so far this year. So ready for some sunshine and warmer days.

 

[wanttaja]

Could be the economy, fewer people flying small aircraft. I always thought the cause was usually pilot error. It sure seems this way if you read the accident reports. Things like losing control under IFR conditions is a big one, running out of fuel is popular, missed approach screw ups, on and on.

Pilot Error is certainly the most common accident cause, though the continued VFR into IFR conditions is only a minor portion of it (though the most deadly). Most of pilot error is just plain stick-and-rudder skills....IIRC, about 38% of homebuilt accidents involve basic flying skills, while about 15% of them involve errors in judgement (VFR in IFR, buzzing, running out of fuel, etc.).

I compared the statistics for homebuilts vs. a control group of Cessna 172s and 210s, and the control group results were higher: About 55% stick-and-rudder vs. ~20% judgement. What was interesting was the median flight hours for the pilots of the homebuilts was about twice that of the control group.

The FAA compares the relative safety of homebuilts vs. production airplanes based on accidents per 100,000 flight hours, and that flight hours prediction ties into the assumption for the annual use that I mentioned in the last post. The FAA says that homebuilts have an accident rate ~500% higher than overall GA.

When I compare them strictly by aircraft fleet size (# of accidents vs. number of registered examples), homebuilts have a 40% higher accident rate. When I leave off planes in their test period, homebuilts are about 17% higher.

But then, I get the "so many registered homebuilts don't fly" and "a guy with a Luscombe flies much more often than someone with a Lancair" arguments.

Sigh.

Ron Wanttaja

 

[wanttaja]

Just from my casual observations, it seems that the number of hours of flying that the average Vans RV owner flies annually, have actually picked up a little bit over the past two years. At least this has been noticeable amongst all the RV'ers I know here in the north TX / southern OK region, and by conversations I've had with the folks I hung out with in HBC at Airventure in 2012 and 2013 (record numbers of RVs showed up in HBC/HBP those years, of course Van's 40th Anniversary in 2012 had a lot to do with that).

I've used the accident statistics to estimate the annual use for homebuilt types. The last time I ran it, RVs, Glasairs, and the like came out about ~80 hours per year. The light fun stuff (Kitfoxes, Fly Babies) seems to be a bit more than half that (~45 hours).

Ron Wanttaja

 

[Geico266]

I have done several first flights on homebuilts. I agree that the first few flights should be done solo by a highly experienced and current pilot. I don't want any distractions from the task and I will not do an initial test flight with a crowd there to watch. I usually have the owner and my good friend who is an IA and experienced in test flying himself present. Usually you can work out any problems in 5 hrs or less. It would then be nice to legally put the owner in the airplane and get him transitioned so he is able to fly off the remaining hours. How many builders want to let/pay someone else to fly off the 40hr test period? More and more builders do let an experienced pilot do the initial testing because now most planes like RVs are in the $100,000 range and they don't want to risk such a significant investment. I think this has cut down on the accident rate in the test period. Don

Interesting take Dan.

A good ground crew taking notes is a good thing to have IMHO. Lessens the work load.

 

[wanttaja]

Well, there were 2,296 total aircraft accidents in 1998, and 1,804 in 2012. So claiming the homebuilt rate has improved is a bit tougher. When I get home tonight, I'll do a fleet-rate comparison between the years.

OK, here we go. I used my own copies of the FAA registration database for the fleet sizes, and the online NTSB database for the number of accidents.

Year 2000:

Total aircraft registered: 343,873
Total accidents: 2,220

Homebuilts registered: 23,286
Homebuilt Accidents: 241

Fleet rate:
Overall: 0.64%
EX-AB: 1.0%

Homebuilts ~54% higher.

Year 2012:
(this reflects some of the deregistrations)

Total aircraft registered: 352,198
Total accidents: 1,804

Homebuilts registered: 32,041
Homebuilt Accidents: 219

Fleet rate:
Overall: 0.51%
EX-AB: 0.68%

Homebuilts ~33% higher.

The 2013 figures are a bit lower... 0.46% for all aircraft, 0.55% for homebuilts. This is with a considerable reduction in fleet sizes, as all the de-registrations seemed to take effect in 2013.

Ron Wanttaja

 

[wanttaja]

This is with a considerable reduction in fleet sizes, as all the de-registrations seemed to take effect in 2013.

BTW, people might be interested in the results of the FAA's re-registration effort. This was a three-year effort that ended on 31 December 2013; if an owner did not re-register his aircraft, the registration was cancelled.

So...how many got cancelled?

As of 31 December 2010, there were:

373,869 total aircraft registered
34,916 were listed as Experimental Amateur-Built

As of 15 November 2013:

317,588 total aircraft registered (About a 15.1% decrease)
27,877 listed as Experimental Amateur-Built (a 14.7% decrease)

The FAA includes a list of deregistered aircraft with the database download.

72,888 total aircraft were deregistered from 1 October 2010 to 15 November 2013

7,672 homebuilts were deregistered over the same period.

Almost 20% of all US aircraft (19.5%) have been removed from the rolls, and about 23.5% (nearly a quarter) of the homebuilts.

Ron Wanttaja

 

[ja_user]

BTW, people might be interested in the results of the FAA's re-registration effort. This was a three-year effort that ended on 31 December 2013; if an owner did not re-register his aircraft, the registration was cancelled.

So...how many got cancelled?

As of 31 December 2010, there were:

373,869 total aircraft registered
34,916 were listed as Experimental Amateur-Built

As of 15 November 2013:

317,588 total aircraft registered (About a 15.1% decrease)
27,877 listed as Experimental Amateur-Built (a 14.7% decrease)

The FAA includes a list of deregistered aircraft with the database download.

72,888 total aircraft were deregistered from 1 October 2010 to 15 November 2013

7,672 homebuilts were deregistered over the same period.

Almost 20% of all US aircraft (19.5%) have been removed from the rolls, and about 23.5% (nearly a quarter) of the homebuilts.

Ron Wanttaja

How many are still trying to get through the paperwork of getting registered?

 

[Jim Logajan]

Might want to inform the FAA. If you fly a 172, they assume you fly 200 hours per year. If you complete your RV-8 and sell your 172, they assume you now fly only 29 hours per year.

One possible statistical cross-check is to take a random sample of airplane "For Sale" ads that contain total hours and year built. One could then compute average hours flown per year and by grouping the airplanes into factory vs experimental, one can see whether the FAA's assumptions are remotely reasonable.

Did they state what kind of statistical sampling or surveys they used to justify their flight hour numbers? Was the same methodogy used for all aircraft groups?

 

[Neal Howard]

I've used the accident statistics to estimate the annual use for homebuilt types. The last time I ran it, RVs, Glasairs, and the like came out about ~80 hours per year. The light fun stuff (Kitfoxes, Fly Babies) seems to be a bit more than half that (~45 hours).

Ron Wanttaja

Around 80 per year seems about right for an average for most of the RVers I know.

Most spamcan owners I know are down to around 50 hours per year or less these past few years, down from a former average of around 100/year before the economy tanked.

The last year I owned my Cherokee, I only put 16 hours on it... but put 70 hours on an RV-8 that me and a couple friends built
...then I bought my RV-6 and put over 100 hours on it the first year of owning it (2012) and will have at least another 80 or more by the time my 2nd anniversary comes up in May.

 

[wanttaja]

One possible statistical cross-check is to take a random sample of airplane "For Sale" ads that contain total hours and year built. One could then compute average hours flown per year and by grouping the airplanes into factory vs experimental, one can see whether the FAA's assumptions are remotely reasonable.

Did they state what kind of statistical sampling or surveys they used to justify their flight hour numbers? Was the same methodogy used for all aircraft groups?

I never got a full explanation for their hour estimate. It was during a telecon between me, EAA, FAA, and AOPA regarding the Nall Report announcement of a big spike in homebuilt accident rate, a few years back. I was just told of the hour assumption, and that it reflected the number of assumed inactive airplanes and the results of their annual survey.

(BTW, for those who don't remember, the "spike" reported by the Nall Report was due to the NTSB's faulty "Amateur-Built" flag in their records. The Nall Report was counting SLSAs and ELSAs (many former ultralights) as Amateur-Built aircraft.)

I run the same sort of hours/year average on the accident data for the production class, but am not sure whether it can be directly compared to the homebuilts. I get about 200 hours per year, but most of those production planes are way old, and many were on the rental lines for years to built thousands of hours.

The planes *I* see are the survivors (or they at least survived long enough to be included in my modern-day lists). How many planes of the same model fell by the wayside in the 40 years since a given example was built? That's like going to a retirement community and claiming that humans live to an average of 88 years old because that's the average age of the folks who die in the facility.

Ron Wanttaja

 

[zaitcev]

The last time I ran it, RVs, Glasairs, and the like came out about ~80 hours per year. The light fun stuff (Kitfoxes, Fly Babies) seems to be a bit more than half that (~45 hours).

It was an unpleasant surprise for me, although I should've foreseen it. I stopped renting and bought a cheap airplane because it was getting too expensive, and therefore I wasn't getting flight hours that I consider minimally necessary. In 2011 I flew 86 hours, in 2012 - 72. So I was thinking that flying on the cheap would let me easily do 100..120 hours, while actually saving money (including the amortized cost of purchase over the expected ownership period and all the other expenses). But in fact I only logged 7.4 hours in 2014, so I'm on track for about 42 hours per year. That is despite flying every weekend! The problem is, I am afraid to take my toy airplane to long cross-countries, so all I do is putzing around the pattern (I made about 50 landings).

 

[DaleB]

The problem is, I am afraid to take my toy airplane to long cross-countries, so all I do is putzing around the pattern (I made about 50 landings).

So, I'm curious. Would you say you bought the wrong plane, need to bite the bullet and get it fixed so you have confidence in it, or just need to build enough time to be confident? Or something else?

 

[zaitcev]

So, I'm curious. Would you say you bought the wrong plane, need to bite the bullet and get it fixed so you have confidence in it, or just need to build enough time to be confident? Or something else?

It's a combination of things. I have some parts on order that should fix the last safety squack that I identified, but beyond that it turned out that I handwaved performance differences too easily. My thought was "well, it's like an old Cub and people fly L.A. to NYC in them". But it's not quite so easy. I cannot take any mountain passes that I took in Remos and Cherokees, and the influence of wind is much exhaggerated now. It's trivial to fly over a ridge and not being able to get back. It imagine all that would matter less if I lived in Florida, but I am in New Mexico. As it is, I'm still going to visit Denver (FTG) for Rocky Mountains Expo in May, but what used to be a 3 hour hop in Cherokee is now my biggest aviation adventure yet, with careful monitoring of weather, multiply fuel stops, and routing away from terrain.

P.S. Actually Denver is not even the worst... I sketched a trip to Phoenix Casa Grande too (they often have fly-ins), and the only way to get there is the long way around down Rio Grande valley and along I-10. It takes at least 10 hours versus 3 hours in Cherokee.

 

[yakdriver]

Pete, What kind of airplane did you get? Don

 

[zaitcev]

Don, there are pictures in the thread "Carlson Sparrow and Sport Special":
http://www.pilotsofamerica.com/forum/showthread.php?t=63772

BTW, I still think this looks like a great value -- it's a an airplane broadly similar to my Sparrow:
http://www.barnstormers.com/classified_868954_Preceptor++N-3+PUP.html
Only $8k!

Anyhow, I seem to be right on the center of Ron's figures: it's either ~70..80 hours in a typical light single, or ~40..50 hours in a light experimental that he mentioned. Statistics suggest that I have very little room for improvement.

 

[yakdriver]

Does that have a 503 Rotax? Probably cruise around 85mph? Jetting the carb correctly will make a huge difference in performance on a two stroke. Also an Ivoprop electric variable pitch prop is a great performance enhancement. Don